Temporal GIS do we go back, or only forward?

Marceau et al.’s paper on temporal topology in GIS Databases outlines the faults with temporal GIS which seems echoes Marceau’s earlier paper on spatial GIS and its faults within social and natural sciences. In both temporal and spatial GIS, as I compare the two papers, the resolution seems to be one of the main issues affecting the accuracy of topology. To clarify, in spatial GIS, higher resolution reflects more data acquired and more accurate spatial topology; while in temporal GIS, higher resolution reflects a higher rate of sampling versus change in the area and more accurate temporal topology. To simplify, when dealing with temporal GIS, it comes down to the sample rate and what is included within the sample and thus, as talked about in spatial scale the “politics of scale”.

I believe that Marceau et al. tried to address the intervals of sampling, however I believe, from everything I have read on scale and scale changes, that Marceau et al.’s approach may be fatally flawed in that it is too simple to transfer to larger areas with greater variability of change. The method once upscaled will produce uncertainties if not greater uncertainties then what was already have within there study area. In essence, it is the “politics of scale”, were the question on how temporal GIS operates or can operate within a software platform, is mired in uncertainty by the data collected over time and the modification by Marceau et al. during the application of the data within a platform setting.

For temporal GIS, it may be impossible to go back in time to map. GIScience may have to start new from the present time working into the future. Therefore, GIScientists will know, now that platforms exist, that certain data sets need to be created that can represent the change rather then extrapolation into the past, which is inherently uncertain without wide-scale identifiers present (i.e. Land survey archeology or the process of digging into the ground to identify past land-uses for topology identifiers).



Tags: , ,

Comments are closed.